IRC log of htmlspeech on 2011-11-03

Timestamps are in UTC.

15:13:00 [RRSAgent]

RRSAgent has joined #htmlspeech

15:13:00 [RRSAgent]

logging to http://www.w3.org/2011/11/03-htmlspeech-irc

15:13:42 [burn]

zakim, I am Daniel_Burnett

15:13:42 [Zakim]

sorry, burn, I do not see a party named 'Daniel_Burnett'

15:14:14 [burn]

Present+ Daniel_Burnett

15:16:36 [smaug]

smaug has joined #htmlspeech

15:17:36 [smaug]

hi

15:28:28 [smaug]

Zakim, smaug is Olli_Pettay

15:28:28 [Zakim]

sorry, smaug, I do not recognize a party named 'smaug'

15:28:52 [smaug]

well, how am I then o_O

15:29:05 [smaug]

s/how/who/

15:31:53 [burn]

burn has changed the topic to: agenda: http://www.w3.org/2005/Incubator/htmlspeech/2011/11/f2f/agenda.html (burn)

15:32:50 [burn]

Chair: Daniel_Burnett,Michael_Bodell

16:01:57 [smaug]

pong

16:06:13 [burn]

zakim, code?

16:06:13 [Zakim]

the conference code is 773324 (tel:+1.617.761.6200 sip:zakim@voip.w3.org), burn

16:06:32 [mbodell]

mbodell has joined #htmlspeech

16:07:03 [Zakim]

Team_(HTMLSpeech)12:00PM has now started

16:07:10 [Zakim]

+tpac

16:07:55 [burn]

trackbot, start telcon

16:08:23 [trackbot]

RRSAgent, make logs public

16:08:27 [trackbot]

Zakim, this will be

16:08:27 [Zakim]

I don't understand 'this will be', trackbot

16:08:29 [trackbot]

Meeting: HTML Speech Incubator Group Teleconference

16:08:31 [trackbot]

Date: 03 November 2011

16:09:01 [Zakim]

+[IPcaller]

16:09:19 [smaug]

Zakim, [IPcaller] is Olli_Pettay

16:09:19 [Zakim]

+Olli_Pettay; got it

16:09:39 [smaug]

Zakim, nick smaug is Olli_Pettay

16:09:39 [Zakim]

ok, smaug, I now associate you with Olli_Pettay

16:17:46 [ddahl_]

ddahl_ has joined #htmlspeech

16:20:13 [Milan]

Milan has joined #HtmlSpeech

16:20:32 [Milan]

Topic: Introductions

16:20:37 [glen]

glen has joined #htmlspeech

16:22:59 [DanD]

DanD has joined #htmlspeech

16:23:22 [nagesh]

nagesh has joined #htmlspeech

16:25:23 [Milan]

Topic: Review recently sent examples

16:26:11 [DanD]

http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Oct/att-0064/speechwepapi_1_.html#introduction

16:26:16 [mbodell]

http://bantha.org/~mbodell/speechxg/example1.html

16:27:57 [Milan]

Michael: Speech Web Search Markup only

16:29:37 [Milan]

Robert: Found addGrammarFrom() is awkward

16:30:16 [Milan]

... really a hint

16:30:33 [Milan]

Glen: True that input has no grammar

16:31:16 [Milan]

Michael: It's a builtin grammar

16:31:34 [Milan]

Robert: What about derviveGrammarFrom

16:31:42 [Milan]

Glen: It's an append grammar

16:32:23 [Milan]

DanD: Option might be a better example

16:33:03 [Milan]

Michael: Text is a grammar

16:34:00 [Milan]

Robert: Assume q is an object from which a grammar can be derived

16:34:09 [smaug]

Nit, <button name="mic" onclick="speechClick()"> is a submit button, so when you click it, the form is submitted. type="button" would fix the problem

16:34:18 [Milan]

DanB: addDerivedGrammar

16:34:42 [Milan]

Debbie: Figgure out semantics first

16:35:38 [Milan]

Robert: AddDerivedGrammarFromID

16:35:56 [Milan]

Glen: Also rename q to 'inputField'

16:37:50 [Milan]

Glen: Also from text input type to date or somethign more contrained

16:38:36 [Milan]

... Need to specify the lack of grammars

16:38:47 [Milan]

... Is this dictation?

16:39:34 [Milan]

Robert: improve example by defaulting to UTF-8

16:40:11 [glen]

Section 5.1: when no grammar specified, defaults to builtin:dictation

16:40:31 [Milan]

Robert: Base 64 encoding is ugly

16:41:12 [Milan]

... to the point where it is unsualbe

16:41:40 [Milan]

Michael: Worried about directly inserting XML due to 8th bit

16:42:43 [Milan]

DanB: Are there already common protocols for inserting strings derived from URLs into local variables?

16:43:16 [Milan]

Glen: Should only be a W3C standard, implmentation is orthoginal

16:44:45 [Milan]

Robert: AddFromString() would be nice:?

16:45:14 [Milan]

Glen: addStringGrammar() and addElementGrammar()

16:45:22 [matt]

matt has joined #htmlspeech

16:45:22 [Milan]

Avery: Perfer longer name because its truer to form

16:45:24 [ehlen]

ehlen has joined #htmlspeech

16:46:09 [smaug]

Couldn't you just prepend "data:application/srgs+xml," to the serialized XML. But anyway, using data urls is kind of hackish, IMO.

16:47:11 [Milan]

Robert: Too many dots to get the interpretation

16:48:15 [Milan]

Milan: Propose addGramamrFromURI()

16:49:25 [Milan]

Robert: Newing up a speech grammar is better approach

16:51:56 [Milan]

Michael: Let's just raise issues now rather than solve them

16:53:02 [Milan]

Debbie: Example is complex, and gets mixed up with arguement that JS is complex

16:53:50 [Milan]

* laptop?

16:55:03 [Milan]

Michael: Next example from Bjorn

16:56:56 [Milan]

Robert: The example lacks a grammar

16:57:11 [smaug]

s/onclick="startSpeech"/onclick="startSpeech(event)"/

16:57:18 [DanD]

http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/att-0008/web-speech-sample-code.html

16:58:45 [Milan]

Robert: Need to define what happens when lacking a grammar

16:59:12 [Milan]

Avery: Is there a policy against comments in the examples?

16:59:39 [Milan]

Michael: Planning on adding that to an appendix

17:00:52 [Milan]

Avery: It's a decent example, as long as it is clear that this instance lacks a grammar

17:01:11 [Milan]

Robert: Example shows default behavior

17:02:03 [Milan]

Rahul: Could also delete button as means of shorting example

17:02:29 [glen]

per Avery's suggestion: add a comment "since no grammar is specified and no element is binded, uses default grammar builtin:dictation"

17:03:28 [burn]

s/adding that to/adding examples to/

17:03:54 [Milan]

Rahul: Two different ways to perform same array access

17:04:02 [Milan]

Glen: Should make it consistent in example

17:06:18 [mbodell]

In Bjorn's second example need sir.maxNBest = 2;

17:06:19 [glen]

use same notation: s/q.value = event.result.item(0).interpretation;/q.value = event.result[0].interpretation;/

17:07:14 [Milan]

Robert: Intent is to get a text transcript of the user's input

17:07:42 [Milan]

... why are we accessing the interpretation instead of tokens?

17:10:24 [Milan]

Milan: Need to bring this up in protocol team

17:11:17 [Milan]

<all agreed> to replace to "utterance" in place of interpretation

17:12:01 [Avery]

Avery has joined #htmlspeech

17:12:40 [Milan]

Milan: Last two comments should apply here as well

17:15:12 [Milan]

Milan: Should we have company-specific references?

17:15:22 [Milan]

Michael: Prefer example.org

17:15:54 [Milan]

Robert: Is there speech recognition in turn by turn>

17:16:17 [Milan]

Michael: Speech recognition is just destination capture

17:16:38 [smaug]

Again, s/onclick="startSpeech"/onclick="startSpeech(event)"/

17:17:14 [Milan]

Robert: The prefer speek next instruction should cancel last instruction

17:17:47 [Milan]

Glen: Thought the purpose of example was to show interplay between speech and tts?

17:18:13 [Milan]

Michael: TTS play resumes where last left off

17:18:41 [Milan]

Glen: Way to stop prior play is a good feature

17:19:15 [Milan]

... we should change this example

17:19:44 [glen]

change example to show how to stop, by persisting the tts object and calling stop before adding .text and .play

17:21:13 [Milan]

Michael: Ollie example next

17:21:25 [mbodell]

http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/att-0009/htmlspeech_permission_example.html

17:24:44 [Milan]

Micahel: First example is just removing unauthorized elements?

17:25:12 [Milan]

... but second example doesn't allow speech input to start

17:25:13 [Milan]

Ollie: Yes

17:25:56 [Milan]

Michael: Can you transition from not authorized to authorized?

17:26:14 [Milan]

Ollie: Should be possible, but example doesn't do that

17:27:13 [Milan]

... but could also just reload the page

17:27:56 [Milan]

* Going on break now

17:30:24 [burn]

s/Topic: Introductions/ScribeNick: Milan/

17:51:11 [ddahl_]

scribe:ddahl

17:51:19 [ddahl_]

topic: Robert's example

17:51:46 [ddahl_]

robert: two recognitions in a row, you want to pick your cities based on what state you're in.

17:52:19 [rahul]

rahul has joined #htmlspeech

17:52:52 [Avery]

Actually I think it's based on what state is specified in the first reco, not necessarily what state you're in. A minor nit.

17:52:56 [ddahl_]

robert: it really should say "interpretation.state", not just "interpretation"

17:53:35 [ddahl_]

...used push instead of adding things to the array of speech grammars

17:55:41 [ddahl_]

robert: a bug on result, should be city, also, sr.onMatch should be sr.onResult

17:56:20 [ddahl_]

robert: second example is rereco

17:57:35 [ddahl_]

...gives grammars to speechInputRequest, then classifies, then does rereco with a specific grammar

17:58:12 [ddahl_]

glenn: this seems to be a strange use of "interpretation"

17:58:27 [ddahl_]

robert: there is a huge universe of grammars

17:58:50 [ddahl_]

rahul: this is identifying one grammar as different from the others

17:59:15 [ddahl_]

robert: using the attribute "modal" to activate and deactivate grammars

17:59:43 [ddahl_]

...would change the example to get interpretation.classification

18:00:27 [ddahl_]

...strange to have multiple "modals" as true, think modal might be a bad idea

18:01:01 [ddahl_]

topic: speech-enabled email

18:01:46 [ddahl_]

michael: one interesting thing is that you might get notifications that you would want to speak to, but without clicking

18:02:22 [ddahl_]

robert: was mostly thinking about things like "reply", but you could also imagine saying "read it to me" after notification

18:02:47 [ddahl_]

...made up a method to cancel TTS

18:03:24 [ddahl_]

michael: you could just delete the element

18:03:36 [ddahl_]

robert: what if you set up the element with stuff in it?

18:03:49 [ddahl_]

glenn: destroy should not be to only way to cancel

18:05:44 [ddahl_]

topic: Milan's example of protocol

18:06:12 [ddahl_]

milan: will augment with API calls that trigger protocols

18:07:34 [glen]

glen has joined #htmlspeech

18:07:46 [ddahl_]

...need a result index of some kind

18:08:12 [ddahl_]

...then recognizer decides to change its mind and reorders results

18:10:26 [ddahl_]

...strange to get a "complete" result in the middle of a long dictation

18:11:29 [ddahl_]

...result index 0 is the first fragment, then halfway through the second fragment, the recognizer says the first one is done

18:11:51 [ddahl_]

...different from MRCP, because in MRCP that means it's the end of it

18:12:41 [ddahl_]

...then retracts a result, not sure how to represent this, maybe an "IN_PRO

18:12:59 [ddahl_]

...GRESS" message with no payload

18:13:30 [ddahl_]

... we will put this in the larger document as an example of the protocol

18:14:02 [ddahl_]

topic: michael johnston's multimodal use case

18:14:19 [smaug]

Could you please paste links to the example here

18:14:21 [ddahl_]

michael: "I want to go from here to there" is the use case

18:14:40 [smaug]

( would be then easier to read minutes later )

18:15:39 [mbodell]

Michael's example: http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/att-0020/multimodal_example.html

18:16:15 [mbodell]

You can walk through the examples from: http://bantha.org/~mbodell/speechxg/f2f.html which links to http://bantha.org/~mbodell/speechxg/examples.html which then walks through the examples

18:16:30 [ddahl_]

glenn: it would be good to have a "state" attribute

18:17:12 [ddahl_]

...the "nomatch" state is more of a result, not a state

18:17:42 [ddahl_]

...we may need more than one attribute to get results of speech processing

18:18:21 [ddahl_]

michael: this also has the EMMA so that you can see the mapping from EMMA

18:18:37 [ddahl_]

...this example makes use of a remote speech service

18:19:11 [glen]

http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/att-0020/multimodal_example.html

18:19:24 [ddahl_]

...the EMMA shows the combined speech and gui input

18:20:13 [ddahl_]

robert: this should be a wss: , that is, a web socket protocol, but what should we do if someone uses http?

18:21:40 [ddahl_]

michael: you could get the command right but not the person if you didn't do the "clickInfo"

18:22:02 [howard]

howard has joined #htmlspeech

18:22:27 [ddahl_]

topic: Charles Hemphill's example

18:22:27 [glen]

http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/0024.html

18:23:00 [ddahl_]

danD: we should start with the simplest example

18:24:21 [ddahl_]

topic: Michael Bodell's example 8, translation

18:24:31 [glen]

http://bantha.org/~mbodell/translate.html

18:24:41 [glen]

view-source:http://bantha.org/~mbodell/translate.html

18:24:42 [ddahl_]

michael: different example of translation

18:25:15 [ddahl_]

...there's from and to languages, you choose, and then click on microphone to talk

18:25:45 [ddahl_]

...there's a progress bar that get's updated

18:26:26 [ddahl_]

...we're grabbing our language from the selector, we're using a dictation grammar for whatever language we're using

18:26:43 [ddahl_]

...where are we doing capture?

18:26:57 [ddahl_]

glen: wouldn't that be the microphone?

18:27:14 [ddahl_]

michael: not necessarily, there could be other things like media streams

18:27:58 [ddahl_]

glen: is capture necessary or does it just provide more features?

18:28:52 [ddahl_]

michael: we didn't have any examples of capture from other places, like from Web RTC

18:29:13 [ddahl_]

...right now there's no standard for accessing microphone

18:29:40 [ddahl_]

glen: would like to see default example where we don't have to explicitly do capture

18:30:11 [ddahl_]

michael: all examples assume that there's magic for capturing audio

18:30:28 [ddahl_]

glen: can't we make it so that the magic is what happens by default?

18:30:50 [ddahl_]

dan: there are many security and privacy issues

18:31:13 [ddahl_]

...different permissions for getting access to media but also to do something to the media

18:32:00 [ddahl_]

michael: this is also raised in some of our issues, we only have a two sentence note now

18:32:22 [ddahl_]

...can TTS work on Web Sockets?

18:32:25 [ddahl_]

robert: yes

18:33:25 [ddahl_]

michael: on audio start, etc. are in our spec. another issue is that payload of start, stop events isn't defined

18:33:56 [ddahl_]

robert:: do we have VU meter events?

18:34:14 [ddahl_]

michael: no

18:34:25 [ddahl_]

dan: that came up in Web RTC, they don't have that, but they could create it

18:34:43 [ddahl_]

michael: we do have speech-x events for custom extensions

18:34:57 [ddahl_]

robert: most speech apps have one

18:35:11 [ddahl_]

michael: is that part of the UA or the app?

18:35:32 [ddahl_]

topic: Debbie's example

18:35:52 [ddahl_]

multi-slot filling

18:35:56 [mbodell]

Debbie's: http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/att-0031/Multi-slotSpeech1.html

18:36:16 [DanD]

DanD has joined #htmlspeech

18:37:43 [ddahl_]

debbie: in this example you have to pull out the slot values from the EMMA

18:38:54 [ddahl_]

robert: is this the same as saying "interpretation.booking"?

18:39:21 [ddahl_]

debbie: not sure

18:39:50 [ddahl_]

... we don't know what's in "interpretation"

18:40:48 [ddahl_]

robert: we could get rid of "interpretation"

18:41:16 [ddahl_]

michael: it could be a useful pointer into the EMMA

18:41:57 [ddahl_]

michael: that is available in VXML

18:42:38 [mbodell]

Issue: we should make sure it is clear what the interpretation points to

18:42:38 [trackbot]

Created ISSUE-1 - We should make sure it is clear what the interpretation points to ; please complete additional details at http://www.w3.org/2005/Incubator/htmlspeech/track/issues/1/edit .

18:43:36 [ddahl_]

michael: should do an if to make sure that you really got a value

18:44:21 [ddahl_]

debbie: could add the EMMA

18:44:57 [ddahl_]

...would there be value in some kind of convenience syntax so that you don't need the full DOM generality to manipulate the EMMA result?

18:44:57 [mbodell]

Charles' example: http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/0033.html

18:45:18 [ddahl_]

topic: another example from Charles Hemphill

18:45:36 [ddahl_]

michael: the same example as before but with an external grammar

18:46:12 [ddahl_]

avery: what's the advantage of having "reco" element as a child under "input"

18:46:54 [ddahl_]

michael: there are two different ways to do the same thing, with "reco" under "child" you don't need an id

18:47:07 [smaug]

<input> element can't have child elements

18:47:22 [ddahl_]

s/"child"/ as a "child" under <input>

18:48:00 [kaz]

kaz has joined #htmlspeech

18:48:12 [ddahl_]

actually, input is a child of reco in the proposal

18:48:37 [smaug]

My comments to example 3 http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Nov/0034.html

18:48:57 [ddahl_]

michael: another example with a real inline grammar so that you don't have to do data uri

18:49:38 [ddahl_]

michael: we would have to define a "grammar" tag

18:49:54 [ddahl_]

robert: we would have to define for browsers how to interpret SRGS

18:50:25 [ddahl_]

avery: like putting script in page vs an external reference

18:50:44 [smaug]

Milan: remember, we're talking about HTML here, not XML

18:50:55 [smaug]

(I assume that was Milan)

18:51:13 [ddahl_]

milan: could we say "as long as this is valid XML ignore it and pass it to us"?

18:51:34 [ddahl_]

robert: why wrap the whole thing with the grammar element?

18:51:57 [ddahl_]

michael: if there's an SRGS 1.1, you wouldn't know what version it was, for example

18:52:30 [ddahl_]

...would like to have inline grammar, if any, be full SRGS with <grammar> element

18:53:22 [ddahl_]

michael: that is the end of the examples

18:55:15 [ddahl_]

rrsagent, format minutes

18:55:15 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html ddahl_

18:55:23 [Milan]

* Good point Ollie

18:55:40 [Zakim]

-Olli_Pettay

18:57:32 [howard]

howard has left #htmlspeech

19:24:19 [rahul]

rahul has left #htmlspeech

19:42:35 [burn]

burn has joined #htmlspeech

20:05:14 [glen]

glen has joined #htmlspeech

20:05:20 [glen]

scribenick: glen

20:05:42 [glen]

topic: issues

20:07:38 [Zakim]

+??P0

20:07:47 [smaug]

Zakim, ??P0 is Olli_Pettay

20:07:48 [Zakim]

+Olli_Pettay; got it

20:08:00 [glen]

burnett: if can't agree, depends on importance. If important, capture different opinions in doc.

20:08:38 [nagesh]

nagesh has joined #htmlspeech

20:08:41 [glen]

...(not required to resolve everything in incubator group)

20:08:46 [mbodell]

First issue to discuss: http://bantha.org/~mbodell/speechxg/issuep1.html

20:09:42 [DanD]

DanD has joined #htmlspeech

20:09:54 [glen]

1. What Content-Type do we want to use on an empty message? Use case was nulling out previous candidate recognition.

20:10:14 [glen]

milan: do we have to specify? can it be assumed?

20:10:36 [glen]

...empty means no payload?

20:11:11 [glen]

robert: protocol doesn't require a body

20:11:34 [glen]

...in which case I don't think it needs a content type. Example getParams

20:11:49 [glen]

michael: what about interim results?

20:12:59 [ehlen]

ehlen has joined #htmlspeech

20:13:41 [glen]

...if no content and not content type then nulls out corresponding result. Example: an interim result gets replaced with no result (e.g. if a <cough> is initially recognized as some text)

20:14:36 [glen]

topic: Protocol Issues

20:14:38 [glen]

2. I am skeptical about changing established MRCP event/method names. I sort of agree that LISTEN is better than RECOGNIZE, but do not think the reasons are good enough to warrant ensuing churn.

20:16:14 [glen]

Robert: Microsoft doesn't care if similar to MRCP, rather that it's compatible with our web sockets protocol

20:16:38 [glen]

burnett: web sockets is just a transport

20:16:55 [glen]

...violates many types of protocol design

20:17:17 [glen]

...if standards track, IETF is a logical place

20:17:33 [glen]

robert: so naming doesn't matter much at this point.

20:17:36 [glen]

all: agree

20:18:16 [glen]

burnett: some talk of using SIP to setup, would have to separate signaling and data...which is one thing wrong with this.

20:18:34 [glen]

robert: this is more to illustrate a point that it can be done

20:19:22 [glen]

burnett: companies could implement today, and may not be completely interoperable (as is often the case on first implementations)

20:20:15 [glen]

michael: we agree, not to change names right now. Names will likely to be re-evaluated in a standards track.

20:21:10 [glen]

...minor syntax issues can be called out as a note in the doc.

20:22:34 [glen]

burnett: when gets into a standards group, they look at requirements and take ideas into consideration, but they consider MANY other factors, e.g. security, that drive

20:22:46 [glen]

3. We need a way to index the recognition results. I suggest using a Result-Index header

20:23:57 [glen]

all: agree to add. if a one-shot recognition, it's only [0] and still optional

20:24:31 [glen]

4. It was awkward to use a RECOGNITION-COMPLETE message presumably with a COMPLETE status during continuous speech. Instead, I used INTERMEDIATE-RESULT with a new Result-Status header set to final.

20:25:12 [glen]

robert: just rename RECOGNITION-COMPLETE as RECOGNITION-RESULT

20:26:15 [glen]

...it's an intermediate, unless it's a final response type.

20:26:57 [glen]

burnett: MRCP has separate status code and completion code

20:28:11 [glen]

Milan: we need a complete flag, not sure it was defined. We haven't stated which status codes correspond to which messages.

20:28:53 [glen]

burnett: in MRCP, status is about communication (like 200 OK). In MRCP, the completion code indicates what happened (e.g. successful reco)

20:29:54 [glen]

robert: so status indicates "sending more", so status should be in-progress for continuous reco case.

20:30:41 [glen]

robert: need request state?

20:31:12 [glen]

burnett: request has been made, has it been completed yet? status is success, illegal method, illegal value, unsupported header

20:31:38 [glen]

robert: reco result, 200 OK, in progress

20:32:12 [glen]

5. Perhaps Source-Time should also be required on final results

20:32:14 [glen]

all: yes, everything's fine, more to come

20:34:16 [glen]

Milan: by time have final result, should know start time.

20:34:33 [glen]

all: agree, require only reco result

20:35:13 [glen]

Milan: could be reco result with type = pending

20:35:33 [glen]

michael: pending implies have already started

20:35:44 [glen]

robert: in progress more accurate

20:35:45 [glen]

all: agree to leave as is

20:35:58 [glen]

6. Wanted to confirm that channel identification is being handled by the WebSocket container

20:36:11 [glen]

robert: handled by web socket

20:36:41 [glen]

...if two separate recos, then two web sockets and two audio streams. (Can have 2 grammars active in one reco)

20:36:53 [glen]

milan: continuous hotword case

20:36:57 [glen]

robert: that's continuous reco

20:37:28 [glen]

...start session with hotword and command-control grammar, all is continuous results

20:37:42 [glen]

michael: hard if change over time

20:37:51 [glen]

...because have to pause to change

20:37:56 [glen]

...so not continuous

20:38:13 [glen]

robert: don't want to transmit audio twice, but with two sessions, you must

20:38:43 [glen]

avery: does emma result specifies which grammar?

20:38:58 [glen]

michael: yes

20:39:14 [kaz]

kaz has joined #htmlspeech

20:39:31 [glen]

7. I noticed that Completion-Cause was missing from Robert's spec example in section 4.2.

20:39:55 [glen]

robert: accidental omission, need to add

20:40:31 [glen]

topic: Web API Issues

20:40:44 [glen]

1. To get the reco result I think i have to write "e.result.item(0).interpretation". This is a lot of dots and an index just to get the top result.

20:41:31 [glen]

robert: I want to write e.interpretation -- because most of the time that's what I want (but still could use the verbose way as well)

20:42:20 [mbodell]

Here is the link to where the event is defined: http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Oct/att-0064/speechwepapi_1_.html#speechinputresult

20:43:29 [glen]

milan: e.result.interpretation

20:43:47 [glen]

michael: can already use e.result[0].interpretation

20:44:54 [glen]

glen: we should change utterance to match

20:45:30 [glen]

all: e.interpretation and e.utterance

20:45:39 [glen]

...agreed

20:46:33 [glen]

2. "utterance" has a couple of different meanings in the doc. It's alternatively the recording of what the person said, or the transcript returned by the recognizer.

20:47:19 [glen]

michael: transcript? text? tokens?

20:48:38 [glen]

text and token are over used and confusing

20:49:54 [glen]

robert: but it is text, so not overloading the concept

20:49:59 [glen]

... (unlike token)

20:50:26 [glen]

burnett: transcript, closest to what's actually happening, laymen get it

20:51:46 [glen]

glen: text is not descriptive: interpretation is text, whereas transcript vs interpretation is clear

20:52:59 [glen]

all: agree: rename utterance to transcript

20:53:47 [glen]

5. The "modal" attribute on SpeechGrammar is unnecessarily restrictive

20:53:56 [glen]

Discussion: There are cases where I'll want to have multiple grammars active, but not all, and not just one. Developers would be better off with a boolean enabled attribute on each grammar. Would be useful to clarify the behavior when there is more than 1 grammar with this set to true (only the first in the list is active?) Is this even useful at all? What is the case for having grammars which aren't active in the reco? Can we change the state of the modal/u

20:54:11 [glen]

robert: less lines of code if just set one to true

20:55:26 [glen]

milan: alternatively, could add/remove from grammars array

20:55:49 [glen]

glen: sending all at once allows caching

20:55:55 [glen]

...of grammars

20:57:02 [glen]

...what about continuous case, can grammars change on the fly

20:57:35 [glen]

michael: we decided to simplify by re-calling .start to change grammars or anything else

20:58:12 [glen]

milan: should have a separate way to preload

20:58:46 [glen]

burnett: voicexml has defineGrammar

20:59:14 [glen]

milan: grammar set object on the SpeechInputRequest

21:00:32 [glen]

...I proposing sets of grammars

21:00:58 [glen]

robert: I'd like it flatter, get rid of enabled/disabled -- just delete -- and don't allow preload

21:02:43 [glen]

michael: already have .open that allows preloading

21:02:54 [mbodell]

See web api: http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Oct/att-0064/speechwepapi_1_.html#dfn-open

21:06:14 [glen]

scribenick: glen

21:06:31 [glen]

burnett: nervous about this, we discussed this for a long time and considered many edge-cases

21:06:36 [howard]

howard has joined #htmlspeech

21:06:51 [glen]

robert: alternative: get rid of modal and enable, and just use a bunch of grammars

21:07:26 [howard]

howard has left #htmlspeech

21:07:55 [glen]

avery: if .open has already been called, .start doesn't call it (.start only calls .open if hasn't been opened yet)

21:09:23 [glen]

burnett: wondering if there are performance advantages in reco engine if can call enable/disable as opposed to calling .open multiple times?

21:10:00 [smaug]

start might need to re-call open if authorizationState has changed from not-authorized to authorized

21:10:11 [glen]

milan: MRCP didn't solve this, why should we?

21:10:24 [glen]

...all good MRCP clients do what you're saying automatically, they automatically check for the deltas

21:10:43 [glen]

robert: this runs at web-scale, distributed

21:11:18 [glen]

...big difference between telephony and web

21:12:26 [glen]

michael: options: eliminate model, keep and define what happens if multiple set to true

21:15:49 [mbodell]

mbodell has joined #htmlspeech

21:16:29 [glen]

glen has joined #htmlspeech

21:16:38 [glen]

avery: easier to add later than remove

21:17:58 [matt]

rrsagent, draft minutes

21:17:58 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html matt

21:18:03 [glen]

agree: eliminate .modal

21:18:13 [glen]

5. The interpretation attribute is completely opaque. That may be necessary given that SISR can pretty much return anything. But it'll need some examples to show how to use it.

21:18:25 [matt]

i|scribe:ddahl|scribe:ddahl_|

21:18:30 [matt]

rrsagent, draft minutes

21:18:30 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html matt

21:18:58 [glen]

burnett: there was support for a flat array of interpretations

21:19:23 [glen]

...I didn't like that, Nuance and their customers didn't like it,

21:20:03 [glen]

debbie: use emma to define layout

21:20:40 [glen]

michael: different reco engines may use emma in different ways

21:21:13 [glen]

...fundamentally, .interpretation points to somewhere in emma, which simplifies (and a corresponding .transcript)

21:22:14 [glen]

all: agree, specify which part of emma holds the interpretation

21:22:46 [glen]

michael: mapped to a DOM object, emma literal or node

21:23:36 [glen]

...like debbie's slot-filling example

21:25:13 [glen]

michael: I will send text for this

21:25:36 [glen]

5. The array of SpeechGrammar objects is too cumbersome

21:26:00 [smaug]

something happened to the audio. it is all just noise

21:26:09 [smaug]

though, getting late here

21:26:13 [Zakim]

-Olli_Pettay

21:26:14 [glen]

Discussion: Robert: The array of SpeechGrammar objects is too cumbersome. In most cases I'd like to write something simple like: mySR.speechGrammars.push("USstates.grxml","majorUScities.grxml","majorInternationalCities.grxml"); But I can't. I have new-up a separate object for each one then add it to the array, even when I don't care about the other attributes. Better to just make it an array of URI strings, and add functions for the edge cases. e.g. void ena

21:27:10 [Zakim]

+[IPcaller]

21:27:24 [smaug]

Zakim, [IPcaller] is Olli_Pettay

21:27:24 [Zakim]

+Olli_Pettay; got it

21:27:30 [glen]

void setWeight(in DOMString grammarUri, in float weight); And yeah, I remember arguing the opposite on the phone call. But that's before I tried writing sample code. Glen: "The uri of a grammar associated with this reco. If unset, this defaults to the default builtin uri." Presumably using the grammar attribute overwrites the default grammar, so if a developer wishes to add a grammar that supplements the default grammar, then this alternative should work: re

21:27:44 [glen]

would add clarity. Michael: If you view source on the web api document you'll see the grammar functions and descriptions are there commented out as I anticipated, and agree, with this comment. We should have both functions and array/collections and this makes the things that Robert and Glen describe much easier/better.

21:29:19 [glen]

michael: grammar spec after ? are hints, before builtin: are required and errors if not supported

21:30:45 [glen]

...example: builtin:contacts may recognize names in smartphone

21:31:35 [glen]

...require built:generic

21:35:14 [glen]

burnett: built:generic means I'll take anything you got: if it's just a date grammar, I'll take it.

21:35:19 [mbodell]

We are talking about http://bantha.org/~mbodell/speechxg/issuew5.html but really more about what happens with no grammar

21:35:34 [ddahl2]

ddahl2 has joined #htmlspeech

21:35:53 [glen]

milan: builtin:generic could respond with failure, builtin:dictation could also respond with failure

21:36:30 [glen]

robert: builtin:generic should be builtin:default

21:36:39 [glen]

...and none specified is builtin:default

21:38:18 [glen]

burnett: what if want to use both default and another grammar

21:38:34 [glen]

glen: then add builtin:default and builtin:foo

21:39:40 [glen]

michael: default is not user default, but service or ua default

21:41:41 [glen]

milan: want a way to record without a grammar

21:42:59 [glen]

michael: we define builtin:default, encourage vendors to implement, and state when none specified, it's on by default. (and when other grammars specifed, it can also be added.

21:44:30 [glen]

michael: I like .addGrammer(url, weight) as a simplification from creating object and then setting it

21:44:55 [glen]

robert: .addGrammarFromUrl(url, weight)

21:45:21 [glen]

... .addGrammarFromElement(element, weight) .addGrammarFromString(string, weight)

21:45:52 [burn]

burn has joined #htmlspeech

21:46:02 [Milan]

Milan has joined #HtmlSpeech

21:46:02 [ehlen]

ehlen has joined #htmlspeech

21:46:24 [glen]

robert: better yet: .addUrlGrammar .addElementGrammar .addStringGrammar

21:47:00 [glen]

... but advantage for objects to be alphabetical order, grouped together in docs

21:47:12 [rahul]

rahul has joined #htmlspeech

21:47:31 [glen]

glen: .addGrammarUrl .addGrammarElement .addGrammarString

21:48:22 [glen]

... remove is a JavaScript array operation

21:49:54 [nagesh]

nagesh has joined #htmlspeech

21:51:34 [glen]

michael: also .addCustomParameter(name, value)

21:53:12 [glen]

all: agree: .addGrammarUrl .addGrammarElement .addGrammarString .addCustomParameter

21:55:38 [burn]

rrsagent, draft minutes

21:55:38 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html burn

21:56:30 [smaug]

I think this is enough for me. I'll read the minutes tomorrow and send comments

21:56:35 [Zakim]

-Olli_Pettay

21:56:42 [smaug]

It is midnight here

22:17:11 [smaug]

dark? it has been dark hear for the last 6 hours

22:17:15 [smaug]

here

22:21:18 [rahul]

scribenick: rahul

22:23:00 [rahul]

Topic: Issue 6

22:23:00 [matt]

rrsagent, draft minutes

22:23:00 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html matt

22:23:01 [mbodell]

Link to the current issue: http://bantha.org/~mbodell/speechxg/issuew6.html

22:23:14 [glen]

6. The names are a bit long.

22:23:25 [glen]

Discussion: e.g. "new SpeechInputRequest()" vs "new SpeechIn()" . e.g. "mySR.speechGrammars.push("foo")" vs "mySR.grammars.push("foo")" . e.g. "resultEMMAXML" vs "EMMAXML" or just "EMMA" (call the other one "EMMAText" ) e.g. "inputWaveformURI" vs "inputURI"

22:26:01 [matt]

zakim, who is on the phone?

22:26:01 [Zakim]

On the phone I see tpac

22:26:10 [rahul]

Milan: how about SpeechRequest instead of SpeechInputRequest?

22:29:58 [rahul]

Robert: SpeechRecognizer?

22:30:12 [rahul]

Milan: AudioSynthesizer?

22:31:30 [rahul]

Glen: SpeechReco?

22:32:03 [Milan]

Milan: AudioSynth

22:32:21 [Milan]

* test

22:33:15 [rahul]

Resolution: We will use SpeechReco instead of SpeechInputRequest

22:33:28 [matt]

-> http://en.wikipedia.org/wiki/Bikeshedding Parkinson's Law of Triviality

22:34:26 [rahul]

Action: Editing team to update to SpeechReco

22:34:26 [trackbot]

Sorry, couldn't find user - Editing

22:35:13 [rahul]

Topic: Issue 7

22:35:33 [rahul]

7. SpeechInputRequest.outputToElement() should be an attribute, perhaps 'forElement'

22:36:25 [matt]

-> http://bantha.org/~mbodell/speechxg/issuew7.html Issue 7

22:40:12 [rahul]

Resolution: Replace outputToElement() function with the outputElement attribute

22:40:41 [rahul]

Topic: Issue 8

22:40:45 [rahul]

8. SpeechInputResult has a getter "item(index)". SpeechInputResultEvent has an array "SpeechInputResult[] results.

22:40:53 [rahul]

Discussion: Can we change both to be collections similar to http://www.w3.org/TR/FileAPI/#dfn-filelist (accessible via [] operator and optionally with a .item() method)?

22:41:28 [kaz]

kaz has joined #htmlspeech

22:42:12 [rahul]

Resolution: Accepted

22:42:12 [rahul]

Topic: Issue 9

22:42:13 [matt]

-> http://bantha.org/~mbodell/speechxg/issuew8.html Issue 8

22:42:21 [matt]

-> http://bantha.org/~mbodell/speechxg/issuew9.html Issue 9

22:42:31 [matt]

s|-> http://bantha.org/~mbodell/speechxg/issuew8.html Issue 8||

22:42:32 [rahul]

9. The <reco> element should probably be a void element with no content on its own

22:42:42 [rahul]

Discussion: Satish: http://dev.w3.org/html5/spec/Overview.html#void-elements. I just noticed this in the for attribute's description, missed it in earlier reads: "If the for attribute is not specified, but the reco element has a recoable element descendant, then the first such descendant in tree order is the reco element's reco control." Is there a benefit to doing this over requiring the 'for' attribute to be set and making reco a void element? Charles: I a

22:42:54 [matt]

i|8. Speech|-> http://bantha.org/~mbodell/speechxg/issuew8.html Issue 8|

22:42:57 [matt]

rrsagent, draft minutes

22:42:57 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html matt

22:43:27 [glen]

resolution: can specify with either descendent or with for= attribute

22:43:28 [rahul]

Resolution: Agreed to leave it as-is using either the for or the descendant pattern

22:43:40 [rahul]

Topic: Issue 10

22:43:46 [matt]

-> http://bantha.org/~mbodell/speechxg/issuew10.html

22:43:52 [rahul]

10. TTS is hard

22:43:53 [matt]

|http://bantha.org/~mbodell/speechxg/issuew8.html||

22:44:19 [matt]

i|10.|-> http://bantha.org/~mbodell/speechxg/issuew10.html Issue 10|

22:44:22 [matt]

rrsagent, draft minutes

22:44:22 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html matt

22:45:04 [rahul]

Discussion: Bjorn: I can't see any easy way to do programmatic TTS. The TTS element is at least missing the attributes @text and @lang. Without those, it's pretty hard to do the very simple use case of generating a string and speaking it. It's possible, but you need to build a whole SSML document. For use cases, see the samples I sent earlier today. Dominic: For TTS, I don't understand where the content to be spoken is supposed to go if it's not specified in

22:46:42 [rahul]

Michael: @lang is not missing since it could be inherited

22:46:54 [rahul]

... there is no @text there

22:50:17 [rahul]

Glen: content within <tts></tts> will show up within older browsers

22:50:35 [mbodell]

Discussion is <tts src="data:text/plain,Hello, world"/> versus <tts value="Hello, world"/> versus something else. Note in JS we could define a function so it is pretty similar, but from Markup a little harder to get the function creating the data uri (probably still possible)

22:56:29 [glen]

72<tts value="fahrenheit">F</tts>

23:01:01 [glen]

michael: tts as a markup may render visually a control (play, stop, etc)

23:01:12 [glen]

...other dom can interact

23:01:32 [glen]

glen: most uses of tts need dynamic control -- that is require javascript

23:01:53 [glen]

michael: because tts inherits from media-element, it requires a src attribute

23:04:38 [glen]

glen: <img alt="text">

23:04:53 [glen]

michael: <tts> is not used as an alternative fallback

23:05:14 [rahul]

Dan: usecase for <tts> element is to facilitate easy generation as part of markup rather than generating script

23:07:02 [burn]

burn has joined #htmlspeech

23:12:38 [rahul]

s/|http://bantha.org/~mbodell/speechxg/issuew8.html||//

23:12:46 [rahul]

rrsagent, draft minutes

23:12:46 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html rahul

23:15:13 [rahul]

Michael: the @lang inherited from the <media> element should be passed as a parameter to the synthesizer

23:15:48 [rahul]

Resolution: Add a @text attribute to <tts>.

23:16:14 [rahul]

Topic: Issue 11

23:16:40 [rahul]

-> http://bantha.org/~mbodell/speechxg/issuew11.html Issue 11

23:16:50 [rahul]

rrsagent, draft minutes

23:16:50 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html rahul

23:17:17 [rahul]

11. How does binding to button work

23:18:12 [rahul]

Discussion: Satish: "When the recoable element is a button then if the button is not disabled, then the result of a speech recognition is to activate the button." "For button controls (submit, image, reset, button) the act of recognition just activates the input." "For type checkbox, the input should be set to a checkedness of true. For type radiobutton, the input should be set to a checkedness of true, and all other inputs in the radio button group must b

23:20:40 [smaug]

smaug has joined #htmlspeech

23:24:50 [rahul]

Michael: propose to have an issue note that this needs further thought

23:25:20 [rahul]

Robert: define what we can, and for others say there is no binding

23:28:02 [rahul]

Resolution: Add issue note that more work to be done on bindings

23:28:21 [rahul]

Topic: Issue 12

23:28:30 [rahul]

12. What about meter, progress, and output elements?

23:28:37 [rahul]

Discussion: Satish: The meter, progress and output elements all seem to be aimed at displaying results and not for taking user input. Is there a reason why these are included as recoable elements?Michael: This is specified at Reco Bindings. A person could want to be able to speak and have it change a progress bar or meter or output element. The primary reason is matching what is done with label. These are all labelable elements and thus ended up as recoable

23:29:36 [rahul]

Glen: suggest we not talk about bindings to these

23:30:11 [rahul]

Dan: we need to decide which ones to leave out, I agree since these are not even <input> elements

23:30:21 [rahul]

Resolution: Remove these from the recoable elements and bindings

23:30:28 [rahul]

Topic: Issue 13

23:30:38 [rahul]

13. grammars and parameters should be collections

23:30:48 [rahul]

Discussion: Satish: Similar to issue 8, SpeechInputRequest attributes 'grammars' and 'parameters' should probably be turned into a collection as well

23:30:56 [rahul]

Resolution: Accepted

23:31:01 [rahul]

Topic: Issue 14

23:31:21 [rahul]

14. rename language to lang

23:31:29 [rahul]

Discussion: SpeechInputRequest.language should probably be changed to 'lang' to match lang attributes.

23:31:42 [rahul]

Resolution: Accepted

23:31:52 [rahul]

Topic: Issue 15

23:31:59 [rahul]

15. rename iterimResults to interimResultsInterval

23:32:13 [rahul]

Discussion: SpeechInputResult.interimResults should probably be renamed to interimResultsInterval to indicate its usage similar to how other attributes have 'Timeout' in their names

23:33:16 [rahul]

rrsagent, draft minutes

23:33:16 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html rahul

23:33:37 [rahul]

Resolution: Turn into boolean property, name does not change

23:33:54 [rahul]

Topic: Issue 16

23:34:02 [rahul]

16. drop enum prefixes

23:34:11 [rahul]

Discussion: SPEECH_AUTHORIZATION_ prefix could be dropped for the enums and just have 'UNKNOWN', 'AUTHORIZED' & 'NOT_AUTHORIZED' (similar to XHR States). Same for SPEECH_INPUT_ERR_* and other such enums.

23:35:52 [rahul]

Resolution: Accepted (given Satish's input and expertise)

23:35:59 [rahul]

Topic: Issue 17

23:36:11 [rahul]

17. A way to uncheck automatically by speech?

23:36:14 [rahul]

Discussion: Glen: "For type checkbox, the input should be set to a checkedness of true." It would be nice to have a way to allow user to say something to set it to false, but I can't think of a good convention for this other than adding an attribute or grammar. Perhaps this could/should only be possible via scripting. (I don't like the idea of toggling the checkbox because some users may not be able to easily observe what state the checkbox is currently in.)

23:36:45 [rahul]

Resolution: See resolution to issue 11

23:37:19 [smaug]

mbodell: I'm kind of online

23:37:23 [smaug]

what enum conflicts?

23:37:35 [smaug]

if the enum is in an interface, then no

23:37:43 [smaug]

s/enum/const/

23:38:13 [rahul]

Topic: Issue 18

23:38:20 [rahul]

18. Binding hints versus requirements

23:38:27 [rahul]

Discussion: Glen: "For date and time types ... type of color ... type of range the assignment is only allowed if it is a valid ..." On our call we discussed how these grammars are hints, and in particular how pattern may be difficult to implement. We discussed that showing an output response, even an invalid one, may be more valuable than no response. Michael: We can do hints for patterns on text, and for numbers out of range, but for other types HTML5 is jus

23:39:11 [rahul]

Resolution: See resolution to issue 11

23:39:12 [glen]

satish provides this example of two sets of enums, with no prefixes.

23:39:21 [glen]

https://developer.mozilla.org/en/DOM/HTMLMediaElement

23:39:27 [rahul]

Topic: Issue 19

23:39:35 [rahul]

19. Does reco and TTS need to be on a server as opposed to client side?

23:39:41 [rahul]

Discussion: Dominic: The spec for both reco and TTS now allow the user to specify a service URL. Could you clarify what the value would be if the developer wishes to use a local (client-side) engine, if available? Some of the spec seems to assume a network speech implementation, but client-side reco and TTS are very much possible and quite desirable for applications that require extremely low latency, like accessibility in particular. Is there any possibility

23:39:53 [matt]

i|18.|->http://bantha.org/~mbodell/speechxg/issuew18.html Issue 18|

23:40:01 [matt]

-> http://bantha.org/~mbodell/speechxg/issuew19.html Issue 19

23:42:59 [glen]

satish continues: HTMLMediaElement.LOADED so no clashes

23:43:45 [glen]

(above refers to issue 16)

23:46:44 [rahul]

Resolution: The service does not need to be remote, UAs may define URIs to local engines. We should add clarifying text specifying this.

23:46:54 [rahul]

Topic: Issue 20

23:47:01 [rahul]

20. Set lastMark?

23:47:11 [rahul]

Discussion: Dominic: An earlier draft had the ability to set lastMark, but now it looks like it's read-only, is that correct? That actually may be easier to implement, because many speech engines don't support seeking to the middle of a speech stream without first synthesizing the whole thing. Michael: Actually the speech xg version has never supported setting a lastMark. You can control playback using the normal media controls (setting currentTime, seekabl

23:47:36 [rahul]

rrsagent, draft minutes

23:47:36 [RRSAgent]

I have made the request to generate http://www.w3.org/2011/11/03-htmlspeech-minutes.html rahul

23:47:45 [burn]

s/specifying this./specifying this. Also, the serviceURI does not need to be remote. We will clarify this as well./

23:53:12 [rahul]

Resolution: Leave as-is right now. Add issue note about also making it writeable.

23:53:15 [rahul]

Topic: Issue 21

23:53:21 [rahul]

21. More frequent callbacks?

23:53:27 [rahul]

Discussion: Dominic: When I posted the initial version of the TTS extension API on the chromium-extensions list, the primary feature request I got from developers was the ability to get sentence, word, and even phoneme-level callbacks, so that got added to the API before we launched it. Having callbacks at ssml markers is great, but many applications require synchronizing closely with the speech, and it seems really cumbersome and wasteful to have to add an s

23:55:42 [rahul]

Resolution: Leave as-is. Suggest as enhancement to SSML.

23:55:58 [rahul]

Topic: Issue 22

23:56:00 [rahul]

22. How do we fit with capture/input/MediaStream?

23:56:11 [rahul]

Discussion: Michael: Our spec has: attribute MediaStream input; but we have nearly no explanation of it and our examples don't show how to use it. Can we do better?

23:57:34 [mbodell]

spec link: http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Oct/att-0064/speechwepapi_1_.html#dfn-input